Adrian, Thank you for your reply.
Do you mean by "you don't allow jsf to clear the tree" - the tree of UI components or the values stored in the UI components that are bound? In the latter case I noticed that bound components in session bean persist their input values until I manually reset them (as described here [1]. Am I understanding you correctly? Regards, Svi [1] http://www.mail-archive.com/[email protected]/msg40105.html 2007/8/28, Adrian Mitev <[EMAIL PROTECTED]>: > Binding to jsf components should be done using request scoped beans only > (with session scope bean you don't allow jsf to clear the tree after the > request is processed). > > 2007/8/27, Svilen Ivanov < [EMAIL PROTECTED]>: > > I'm trying to cluster a certain MyFaces ( 1.1.5, tomahawk 1.1.6) > > application in Tomcat 5.5.20. The applications heavily relies on > > session beans which use ListDataModel and HtmlDataScroller as > > properties. When I setup Tomcat to replicate sessions I got > > NotSerializableException for HtmlDataScroller. > > > > As far as I understood from MyFaces FAQ [2] list data model is not > > serializable and may be marked "transient". Some guys recommend to > > build new data model out of a list [1] each time data model is > > requested. However I couldn't find how to solve the problem with > > HtmlDataScroller. > > > > My initial idea is to subclass MyFaces classes I use in the session > > beans and make them implement Serializable interface. How does this > > sound? > > > > Also I have bindings of input controls (HtmlInputText, > > HtmlSelectOneRadio) in these session beans. Are they also stateless as > > ListDataModel? > > > > I would really appreciate if you share your experience with clustering > > MyFaces application as well or if you point me towards resources on > > the Internet that might help > > > > Regards, > > Svilen > > > > [1] > http://www.mail-archive.com/[email protected]/msg38166.html > > [2] http://wiki.apache.org/myfaces/FAQ > > > > -- > > Svilen Ivanov > > http://svilen-online.blogspot.com > > > > There is no dark side of the moon really. > > Matter of fact it's all dark. > > > > -- Svilen Ivanov http://svilen-online.blogspot.com There is no dark side of the moon really. Matter of fact it's all dark.
